Transmission, Diagnosis, and Recommendations for Control of Pseudoloma neurophilia Infections in Laboratory Zebrafish (Danio rerio) Facilities

Abstract

The microsporidium Pseudoloma neurophilia represents a considerable challenge for laboratory zebrafish (Danio rerio) facilities. In 2010, P. neurophilia infections were diagnosed in zebrafish from 74% of the facilities that submitted fish to the Zebrafish International Resource Center (ZIRC) pathology service, and this organism remains the most commonly diagnosed pathogen in submitted fish. Accordingly, many of the ZIRC pathology service consultations deal with control and prevention of microsporidiosis. Here we describe observations and experiments performed at the ZIRC elucidating aspects of P. neurophilia transmission in zebrafish colonies. We then review current knowledge about P. neurophilia transmission and diagnosis. Considering this information, we present recommendations for control of P. neurophilia in zebrafish facilities.

The use of zebrafish (Danio rerio) as a model organism in biomedical studies of development, genetics, pathology, and small-molecule screens has expanded rapidly. Unfortunately, as the use of zebrafish in research laboratories has increased, so has the prevalence of Pseudoloma neurophilia (microsporidiosis) in laboratory fish populations. First reported in 1980,⁶ the parasite was designated Pseudoloma neurophilia, a new genus and species, in 2001.²¹ In 2010, P. neurophilia was identified in 74% of facilities that submitted fish to the pathology service at the Zebrafish International Resource Center (ZIRC), up from 51% in 2009. The inflammation and structural changes resulting from P. neurophilia infection could affect anatomic, physiologic, and behavior data gathered from infected zebrafish.¹⁵ In combination with stress, the infection also can affect fecundity.²⁶ Therefore, the increased prevalence of this parasite has negative implications for zebrafish husbandry, colony health, and research objectives.

In the past, microsporidian infections in zebrafish were correlated with clinical signs of emaciation and spinal curvature, giving rise to the designation ‘skinny disease.’¹⁶ However, emaciation and spinal curvature are nonspecific signs with numerous potential etiologies including nutritional deficiencies, environmental and water problems, genetic factors, physical trauma, toxins, and various infectious agents.^{2,9,17,18,23,30} Furthermore, we have diagnosed heavy P. neurophilia infections in zebrafish that appeared clinically normal. Therefore, clinical signs alone are an insufficient basis for the diagnosis of microsporidiosis.

At ZIRC, we combine histology with hematoxylin and eosin or Luna staining with a PCR-based assay to detect microsporidia in zebrafish tissues. In the current study we use these assays to elucidate aspects of transmission of P. neurophilia, including potential contamination of male gametes. We then review transmission and diagnosis of the pathogen and suggest strategies to control microsporidiosis in zebrafish facilities.

Materials and Methods

The University of Oregon Institutional Animal Care and Use Committee approved all standard protocols for fish maintenance,³¹ as well as the tests and number of animals described in this study. The University of Oregon's animal care and use program, which includes ZIRC, is AAALAC-accredited.

ZIRC fish facility, husbandry, and management.

Water system.

Zebrafish were maintained in the ZIRC main fish facility, which operates on recirculating water systems with bead (mechanic) and fluidized sand (biologic) filters. Postfiltration UV sterilizers (COM6390, Emperor Aquatics, Pottstown, PA) deliver an irradiance of 45,000 to 70,000 μWsec/cm², assuming a minimum UV transmittance of 90%. Irradiance decreases as lamp power declines from 100% to 80% over its use. It also varies with flow rate through the sterilizer, which fluctuates between 150 and 200 gal/min, depending on the number of tanks in use. City water treated by reverse osmosis is added to replace approximately 10% per day. Salt (Instant Ocean, Spectrum Brands, Madison, WI) is added to a conductivity of 500 μS. Water temperature is maintained at 28.5 °C and pH between 7.2 and 7.6. Separate recirculating water systems feed each half (sides A and B) of the ZIRC main fish room. Water between the 2 systems is separate but can be mixed if needed. All racks on each side of the room receive the same postfiltration and UV-sterilized water. Water is not shared between tanks. Effluent from each tank is plumbed back to the ‘dirty water’ sump to enter the filtration system. A diagram of the ZIRC recirculating water system can be viewed in the Sentinel Fish Program protocol available online (http://zebrafish.org/zirc/documents/protocols.php). Sentinel results and pathogen status are the same for both systems. The nursery and grow-out rack are on side B. When fish are transferred out of the nursery or when tanks are moved or consolidated, fish may move from one side of the room to the other. The health of fish colonies and water-quality parameters on both sides of the room are equivalent.

Rearing embryos, larvae, and adult zebrafish.

Tübingen (TU), AB, and TAB14 (a TU × AB hybrid line) wildtype zebrafish were used in experiments described here. Only surface-sanitized zebrafish embryos enter the ZIRC main fish facility. Embryos and larvae are maintained in glass or culture dishes at a density of 50 individuals per dish in 0.5× E2 medium with 0.5 mg/L methylene blue.²⁴ The ZIRC embryo-bleaching (surface-sanitizing) protocol and recipe for E2 are available online (http://zebrafish.org/zirc/documents/protocols.php). At 5 d postfertilization (dpf), larvae are transferred from culture dishes to static water tanks in the nursery, where they are fed paramecia in 40 mL water twice daily until the fish are 10 dpf. At 10 dpf, the water flow in the nursery tanks is turned on, and paramecia are replaced with dry powder food. The larval fish dry food consists of 2 diets (100- to 150-µm and 150 to 250-µm Larval Diets, Zeigler Brothers, Gardner, PA) mixed in equal proportions. At 12 dpf, 2 feedings of Artemia nauplii (Platinum Grade SF Bay Brine Shrimp, Artemia International, Fairfax, TX) are added to the diet. At approximately 25 dpf, the fish are considered juvenile and moved from the nursery into 1-gal plastic or 20-gal glass tanks. Juvenile fish are fed Great Salt Lake Artemia nauplii (Artemia International) in the morning and commercial flake food at midday and in the afternoon. The juvenile flake food mix consists of 500 g each of 2 larval diets (100- to 150-µm and 150- to 250-µm Larval Diets, Zeigler Brothers) and 250 g Golden Pearls (200- to 300-µm, Artemia International). Fish are considered adults at 3 mo of age. Adult fish are fed Great Salt Lake Artemia nauplii (Artemia International) in the morning and commercial flake food in the afternoon. The adult flake food is a mix of 2000 g Zebrafish Feed (SFF1 Aquaneering, Aquaneering, Inc., San Diego, CA), 200 g Cyclopeeze (Argent Laboratories, Redmond, WA), 500 g Spirulina Flake (OSI, Snowville, UT), and 225 g Golden Pearls (300- to 500-µm, Artemia International). Zebrafish are euthanized by rapid cooling in ice water,³⁴ a method that was approved by the University of Oregon Institutional Animal Care and Use Committee and the NIH Office of Laboratory Animal Welfare.

Tank management and cleaning.

Tanks are cleaned when loose debris begins to accumulate on the bottom of the tank or when algae appear on the front or sides of a tank. Glass 20-gal aquaria are cleaned with the fish in the tank. Tank surfaces are scrubbed (Scotch-Brite General Purpose Scouring Pads, 3M, St Paul, MN). Loosened tank debris is removed either by siphoning or by increasing the water flow to flush the debris out the drain. When 1-gal tanks require cleaning, the fish are transferred to a new clean tank. The dirty tank then is scrubbed (Fine Acrylic Scrubber Pad, Lee's Aquarium and Pet Products, San Marco, CA). Empty scrubbed 20-gal and 1-gal tanks are soaked in 0.7% bleach solution for at least 1 h and then soaked for at least 15 min in 3.175 g/L sodium thiosulfate (Western Chemical, Ferndale, WA) to neutralize the bleach. Tanks are rinsed in 82 °C water in a tunnel washer (Hobart, Troy, OH) and air-dried. Clean, unused nets and scrubbers are used on every tank regardless of tank size or fish population. Nets and scrubbers are rinsed and autoclaved between uses. Personnel typically wear new shoulder-length disposable plastic gloves (Vet-R-Sem, Jorgensen Laboratories, Loveland, CO) when cleaning each 20-gal tank. If gloves are not worn, then hands and arms must be washed with antimicrobial hand soap (BacDown, Fisher Scientific, Pittsburgh, PA) and dried between tanks. Due to the large population size, wildtype stocks typically are reared in 20-gal tanks. These tanks are stocked with 150 to 250 fish. Mutant and transgenic lines may be reared in 20- or 1-gal tanks, with as many as 25 fish in a 1-gal tank. Water flow is adjusted so that approximately 3 to 5 tank-volumes are exchanged every hour.

Fish health monitoring.

Zebrafish are monitored daily for morbidity and mortality. The 1-gal sentinel tanks are stocked with 3 to 4 mo AB wildtype zebrafish and set up in locations that are before and after filtration and UV treatment. Two tanks are maintained at each location, representing exposure durations of 3 and 6 mo. Sentinel fish are sampled quarterly for histopathology. The complete history of ZIRC sentinel results is available online (http://zebrafish.org/zirc/documents/health_report.php). P. neurophilia infections have been detected in prefiltration sentinel animals. Mycobacterium infections are diagnosed occasionally in pre- and postfiltration sentinels. The Mycobacterium species in fish and biofilms at ZIRC was previously identified as M. chelonae through sequencing of the hsp65 gene.³³ The vast majority of Mycobacterium infections at the ZIRC are asymptomatic, and diagnosis is made by histologic sectioning and acid-fast staining.²²

Histology.

Adult zebrafish were euthanized and then fixed in Dietrich fixative for a minimum of 24 h. Dietrich fixative contains 30 mL ethanol (95%), 10 mL formalin (F1635, Formaldehyde 37% solution with 10% to 15% methanol, Sigma-Aldrich, St Louis, MO), 2 mL glacial acetic acid, and 58 mL distilled water. Fixed fish were decalcified in 10% trifluoroacetic acid and processed for paraffin sectioning according to standard techniques. Paraffin sections were stained with hematoxylin and eosin, Luna,^19,25 or Ziehl–Neelsen acid-fast stain.¹¹

A PCR assay for P. neurophilia detection.

A PCR reaction using 2 primer sets was designed to amplify a portion of P. neurophilia DNA and a zebrafish gene as an internal control. Forward primer PNA_03 (5′ TGA AAT GTG GTG ACC CGT TTA GG 3′) and reverse primer PNA_04 (5′ TCC TTG ACC CAT CCT TCC TGT G 3′) amplify a 441-bp portion of the P. neurophilia rRNA gene.²¹ Forward primer POA05 (5′ GCG TCT AGC TTT GCC CTT TGA TG 3′) and reverse primer POA06 (5′ CCG TTT TTG AAG ACA TCT GGT CG 3′) amplify a 186-bp fragment of the zebrafish pou5f1 gene.⁴ To prepare PCR samples, zebrafish tissues were digested overnight at 55 °C in lysis buffer (10 mM Tris-HCL pH 8.0, 50 mM KCl, 0.3% Tween20, 0.3% NP40, 4 mM EDTA) containing 2 mg/mL proteinase K. Digested samples were diluted 1:20 with sterile water. PCR reactions (25 μL) contained 5 μL diluted sample, 0.5 μM PNA_03 primer, 0.5 μM PNA_04 primer, 0.125 μM POA05 primer, 0.125 μM POA06 primer, and 0.3 mM each dNTP. Known P. neurophilia infected and noninfected zebrafish control samples were run with every set of reactions as positive and negative controls, respectively. Cycling parameters for PCR (PTC100 and PTC200 Thermocyclers, MJ Research, Waltham, MA) were 94 °C for 3 min, followed by 40 cycles of 94 °C for 30 s, 60 °C for 40 s, and 72 °C for 40 s, and ending with 72 °C for 5 min, with a hold at 8 °C.

Tissue extraction for monitoring P. neurophilia infections by PCR.

Brain, spinal cord, sperm samples collected from the urogenital opening after gentle abdominal compression (‘stripped sperm’), and dissected testes were analyzed for the presence of P. neurophilia by using the PCR assay described earlier. Neural tissues were dissected from frozen tissue. Adult zebrafish were euthanized, and the carcass was frozen at −20 °C for at least 1 h. By using fine-tip forceps, the top of the skull was detached and the brain removed. To access the spinal cord, fine-tip forceps were used to remove the skin by grasping it at the edge of the empty brain cavity and peeling it back along the body axis. The forceps then were used to remove the skeletal muscle overlying the vertebral column. By using 2 pairs of fine-tip forceps to grasp adjacent vertebrae, the vertebrae were pulled apart to reveal the spinal cord. Between fish, forceps were scrubbed with antimicrobial soap (BacDown, Fisher Scientific), rinsed, dipped in 95% ethanol, and heat-sterilized.

Stripped sperm samples were obtained from adult male zebrafish as described previously.⁵ For this procedure, fish were anesthetized by using tricaine methanesulfonate (Argent Chemical Laboratories) according to a published protocol.⁵ After the procedure, fish were euthanized and positioned in lateral recumbency in a culture dish. Testis dissection was performed by making 2 transverse cuts with microdissecting scissors from the ventral aspect of the fish up to the spine. Cuts were made behind the operculums and just forward of the anal fin. A dorsal plane incision then was made through skin, muscle, and ribs to the coelomic cavity, connecting the dorsal ends of the 2 transverse cuts on one side of the fish. This manipulation creates a tissue flap, which was reflected ventrally to reveal the coelomic cavity. The swim bladder and gut were reflected posteriorly to reveal both testes, which were removed with fine-tip forceps.

Seeding 1-gal tanks with debris from 20-gal tanks.

The 1-gal tanks were stocked with TU fish (age, 100 dpf). Once weekly, one-third of 1 of the 20-gal tanks was scrubbed, and the debris was siphoned from the tank and put into a 1-gal tank of interest. Water flow was stopped for 4 h after the debris was added to the 1-gal tank to ensure adequate exposure time. Of the 4 analyzed 1-gal tanks, one was seeded with debris siphoned from a 20-gal tank containing fish positive for microsporidiosis, one was seeded with debris siphoned from a 20-gal tank with fish negative for microsporidiosis, one received an equivalent volume of clean fish water, and one was maintained without exposure to extraneous debris or water (no action). After 7 wk, the fish in these 1-gal tanks were euthanized, fixed, and processed for histologic sectioning and hematoxylin and eosin staining.

Rearing and sampling of a large uniform population of wildtype TU fish.

A large stock of TU zebrafish was generated for the purpose of studying the occurrence of asymptomatic diseases in wildtype stocks. 750 TU embryos, derived from 2 simultaneous group spawns of 20 to 25 pairs of TU adults, were surface-sanitized, reared in the nursery, and then moved to 1-gal tanks. At 75 dpf, the stock was evenly split among 3 20-gal tanks. These fish were mixed prior to surface sanitization, when they went into nursery tanks, and when they were split into the 20-gal tanks. All tanks were fed and cleaned according to standard ZIRC protocols. Adult fish were sampled from the 20-gal tanks for diagnostic testing. Fish were not removed from the tanks for breeding, shipping, or any other purpose.

Results

PCR for diagnosis of P. neurophilia infections in zebrafish tissues.

A PCR test for P. neurophilia that uses the small subunit ribosomal RNA gene has been developed previously.³² We used a new PCR-based assay, which amplifies a different sequence of the same gene. In addition, our assay uses a second primer set that specifically amplifies the zebrafish pou5f1 gene. This amplification serves as an internal positive control to demonstrate that lack of P. neurophilia-specific product does not result from the failed PCR reaction but from the absence of P. neurophilia genomic DNA in the analyzed sample. Whereas the previous small-subunit PCR assay was tested on known quantities of P. neurophilia spores to assess the sensitivity of the assay,³² we did not statistically analyze our modified protocol here. However, to determine reliability of our PCR assay, we compared PCR results to those obtained from histology. Histology is the most common and informative test used by the ZIRC diagnostic service on inhouse fish and samples submitted by the research community. First, we performed histology and PCR on 2 separate groups of adult TU zebrafish from the same tank. At 369 dpf, 21 fish were processed for histologic sectioning and staining. Of the 21 fish, 12 were positive for P. neurophilia infection by histology. At 400 dpf, another 21 fish from the same tank were euthanized, dissected, and brain and spinal cord digested for PCR. Of these fish, 20 were positive for P. neurophilia infection by PCR.

We then sampled another 20 TU fish at 422 dpf from the same tank and performed PCR and histology on different tissues from the same fish. In this group, 17 brains were positive for P. neurophilia by PCR, and 16 spinal cords were positive by histology. Two fish had negative brain tissue by PCR and positive spinal cord by histology. Three fish had positive brain tissue by PCR and negative spinal cord by histology.

P. neurophilia in sperm samples from infected male fish.

We used the PCR-based assay for P. neurophilia to test whether male zebrafish contribute to the infective material present during spawning. Stripped sperm, dissected testes, brain, and spinal cord from 25 TU male zebrafish were tested for P. neurophilia infection by PCR. Half of the testes were rinsed in water before being placed on the 96-well plate for proteinase K digestion, and half were dissected straight onto the plate. Of the 25 fish, 24 were positive for P. neurophilia in spinal cord or brain or both. However, all sperm and testes from the same 24 fish were negative for P. neurophilia. We repeated this experiment on a second group of 30 AB male fish from a stock previously diagnosed with microsporidiosis. Dissected testes were placed directly on the 96-well plate for digestion without being rinsed in water. By PCR, 28 of the 30 males were positive for microsporidiosis in neural tissue. In addition, positive PCR results were detected in 18 testes and 3 squeezed sperm samples from these male fish.

An additional 27 male fish from this stock were fixed and processed for histologic sectioning. We have not previously observed P. neurophilia spores in histologic sections of testes. However, the anatomy of the testis and the size as well as dark hematoxylin staining of spermatids may confound identification of low numbers of spores. With the Luna stain, microsporidian spores are stained red and are easily differentiated from the fish tissue with blue-stained nuclei.²⁵ The Luna stain was applied to the histologic sections of the 27 male zebrafish to examine the testes for spores. Luna staining of these sections revealed P. neurophilia spores in the nervous system of 26 male fish. In contrast, spores were not seen in the testes of any of these fish. However, Luna-positive spores were identified in inflamed tissue at the swim bladder bifurcation adjacent to the testis in 2 of the fish (Figure 1 A). In another fish, Luna-positive spores were present in the caudal coelomic lining and ventral belly muscle (Figure 1 B and C).

Figure 1.

Luna stain labels microsporidian spores red in histologic sections of zebrafish tissue. (A) P. neurophilia spores in connective tissue dorsal to the testis. (B, C) Spores in ventral skeletal muscle adjacent to the testis. Anterior is to the left, and dorsal is up. Scale bar: 20 μm (A and C), 100 μm (B).

Analysis of tank debris for the potential to transmit infection.

Large (for example, 20-gal) tanks often are cleaned with the fish in the tank by scrubbing the walls and removing debris by siphoning or increasing the water flow to flush debris out the drain. During this process, debris gets mixed in the water column, and fish can swim through and feed on the debris. To determine whether exposure to tank debris increases the risk of contracting microsporidiosis, we seeded a noninfected tank of zebrafish with debris from a tank with microsporidia-positive fish. We set up 4 1-gal tanks each with 20 100-d-old TU zebrafish. At the same time, an additional 16 TU fish from the same stock were processed for histologic sectioning and staining. All 16 fish were negative for microsporidiosis. We identified a 20-gal tank of 1-y-old TAB14 zebrafish as microsporidia-positive by histologic sectioning and staining. Another 20-gal tank of TU zebrafish was tested by histology and identified as negative for microsporidiosis. Of the 4 analyzed 1-gal tanks, one was seeded with debris siphoned from the 20-gal tank containing TAB14 fish positive for microsporidiosis, one was seeded with debris siphoned from the 20-gal tank containing TU fish negative for microsporidiosis, one received an equivalent volume of clean fish water, and one was maintained without exposure to extraneous debris or water (no action). After 7 wk of exposure, 3 of the TU fish exposed to the TAB14 tank debris had developed neural microsporidiosis. Fish exposed to tank debris from the noninfected 20-gal tank of TU fish and controls were all negative for the parasite.

Sampling of asymptomatic fish from a large wildtype population.

To monitor for asymptomatic disease, we sampled TU fish (age, 370 dpf ) from the same breeding event that had been reared in 3 separate 20-gal tanks. To screen the population for mycobacteriosis, 30 fish from each tank were euthanized, fixed, and processed for histologic sectioning and acid-fast staining. As an incidental finding, we noted that 23 of the 30 fish sampled from one tank were infected with P. neurophilia and that all 60 fish from the other 2 tanks were negative for P. neurophilia. At 517 dpf, another 20 fish from each tank were euthanized, and brains and spinal cords were dissected and digested for PCR. From this sample, 18 fish from the tank previously diagnosed with microsporidiosis by histology were positive for P. neurophilia infection by PCR. All 40 fish from the other 2 tanks were negative for the infection.

Discussion

P. neurophilia is an obligate intracellular parasite. Transmission occurs through ingestion of the mature infective spore, which can survive outside the host. Spores are approximately 5.4 × 2.7 μm in size.²¹ Ingestion of spores probably occurs when infected fish are cannibalized;¹³ other evidence supports transmission during spawning.^13,14,27 Foci of chronic inflammation with P. neurophilia spores are not uncommon in the ovary (Figure 2 A). Spores in the ovary likely are released along with eggs during spawning and might then be ingested by adults in the tank or hatching larvae. In addition, spores have been observed in developing eggs in the ovary.¹³ However, whether infected eggs mature and become viable, resulting in potential vertical transmission of P. neurophilia, is not yet known. We noted positive PCR results for P. neurophilia in stripped sperm and dissected testes of zebrafish. However, histologic sectioning and Luna staining did not reveal spores within the testicular parenchyma. Although testes may be infected, results from Luna staining suggest that the positive PCR results on testes and sperm from this stock of males may have been a result of contamination of the samples from nearby tissues. Spores from these locations may inadvertently be mixed with testis or sperm during dissection or stripping. These data indicate that male zebrafish infected with P. neurophilia may contribute to the infective material present at spawning. Presumably male gametes contaminated with spores could be a source of infection for embryos or larvae derived from in vitro fertilization as well, although this association has not been tested directly.

Figure 2.

Hematoxylin and eosin stain reveals P. neurophilia spores in histologic sections of infected zebrafish tissues. Spores are stained light blue with a dark blue sporoplasm. Arrows indicate P. neurophilia spores in (A) the ovary, (B) an area of meningeal inflammation ventral to the spinal cord, (C) a vertebral body, and (D) a ventral nerve extending from the spinal cord. Anterior is to the left, and dorsal is up. Scale bar, 50 μm.

We have documented P. neurophilia infections in 3 successive generations of fish at ZIRC, further supporting the idea that transmission may have occurred during spawning. In addition, zebrafish facilities, including ZIRC, that have been established with surface-sanitized embryos only now have fish colonies positive for the microsporidium. Therefore, the establishment of a zebrafish facility SPF for P. neurophilia was based on screening brood stock at 10 dpf for the presence of infective material.¹⁴

Another microsporidium, Pleistophora hyphessobryconis, the etiologic agent responsible for ‘neon tetra disease,’ has been diagnosed in 3 zebrafish facilities.²⁸ The diagnosis of this parasite has been rare enough that it does not have the same impact on zebrafish husbandry, health, and research objectives as does P. neurophilia. However, many of the strategies recommended to control P. neurophilia also apply to P. hyphessobryconis. Nevertheless, in this manuscript, we focus specifically on P. neurophilia infections.

At this time, a nonlethal test is not available that has sufficient sensitivity to detect microsporidiosis in live zebrafish. Therefore, in addition to assessment of moribund fish, postmortem diagnostic tests should examine fish that are representative of the facility population.

Microsporidian spores, prespore stages, and secondary tissue pathologies can be observed in hematoxylin-and-eosin–stained histologic sections of infected fish (Figure 2). Classic loci of infection are the hindbrain and spinal cord. Adjacent and surrounding tissues, including nerve roots, nerves, vertebrae, and meninges, may also be infected. The microsporidium can be observed in other tissues including skeletal muscle, kidney, ovary, and esophagus. The inflammatory response in infected tissues ranges from mild to severe and granulomatous.

Histology is currently the most informative diagnostic test available in zebrafish for diagnosis of multiple pathologies from a single tissue sample. ZIRC recommends routine sampling of sentinel animals for histologic analysis. We also recommend histology as the initial diagnostic choice for surveillance of moribund fish. Even if microsporidiosis is suspected, histopathology allows evaluation of degree of infection, affected tissues, and alternate pathologies and pathogens.

We compared a PCR-based assay to histologic sectioning with hematoxylin and eosin staining to validate the utility of this assay in diagnosing P. neurophilia infections. Equal numbers of fish from an infected population were examined by both methods, and a greater number of infected fish were detected by PCR. However 1 mo elapsed between the histology and PCR assays, and the infection could have spread within the population during that time. Data gathered by comparing PCR on brain tissue to histologic sectioning and hematoxylin and eosin staining of spinal cords from the same fish highlight the importance of examining both brain and spinal cord, regardless of which diagnostic test is used. When applied to a 96-well plate format, PCR can be used to efficiently screen large numbers of fish. However, because the assay is specific to only a single pathogen, it is best used as a follow-up to histology when monitoring the general health of a population. PCR analysis should be done on brain and spinal cord tissues dissected from frozen samples or freshly euthanized fish. Although this assay can be applied to various tissue samples, neural tissues are frequently the only sites of infection. Furthermore, we have rarely diagnosed P. neurophilia in secondary tissues without also finding the organism in neural tissue. A previously described PCR assay for P. neurophilia has successfully been applied to larval zebrafish samples.^14,32 The assay described in the current study could potentially be applied to other samples as well, including tank debris and embryos, although the digestion step would likely need to be optimized for these samples. Neural tissues contain prespore stages along with mature spores. The DNA in prespore stages may be exposed readily by proteinase K digestion whereas the mature spore, which is the only stage that is stable in the external environment, may require more rigorous means of digestion.²⁷

Examining wet-mount ‘squash’ preparations of dissected brain and spinal cord tissues under light microscopy is a rapid and easy method to screen for microsporidia that does not require sectioning or staining.²¹ However, false negatives may occur due to the focal nature of the infection. One method using Fungi-Fluor stain (Polysciences, Warrington, PA) can be applied to fresh tissues or paraffin sections. It produces fluorescence in the presence of the microsporidian chitinous spore wall.¹³ We have found that Fungi-Fluor stain is particularly useful as a follow-up to hematoxylin and eosin staining, to highlight single or rare spores in nonneural tissues and in areas of inflammation where microsporidiosis is suspected but spores are not easily observed in sections stained with hematoxylin and eosin. Alternatively, Luna stain can be applied to paraffin sections,²⁵ highlighting single spores and those that may be obscured by inflammation or other tissue changes in a section stained with hematoxylin and eosin. Unlike Fungi-Fluor, Luna stain does not require fluorescence, thus allowing better visualization of surrounding tissue structures. Additional stains to detect microsporidian parasites in other hosts and tissues have been described previously and could potentially be applied to zebrafish tissue for P. neurophilia diagnosis.^7,12

We offer 6 recommendations to control P. neurophilia infections in zebrafish populations. First, all imported fish should be reared and bred in a quarantine facility that is physically separate from the main fish facility and that operates on a separate flow-through water system. After spawning, all collected embryos should be surface-sanitized with bleach before being introduced into the main fish facility. Although the concentration of bleach used to sanitize embryos (25 to 50 ppm for 10 min) is insufficient to kill all microsporidian spores,⁸ this practice will prevent the introduction of other pathogens and bleaching along with agitation may decrease the likelihood of spores sticking to the chorions. Moving fish that appear clinically normal from quarantine to the main fish facility is not recommended. Fish may appear physically and behaviorally normal while infected with P. neurophilia, along with other pathogens.

Second, maintaining effective UV sterilization in the filtration system is perhaps the most important means of preventing the spread of microsporidiosis throughout an entire facility. UV sterilization units positioned at the end of a recirculating filtration system will kill microsporidian spores, and other pathogens, and prevent the spread to postfiltration sentinel fish and the rest of the population. Most UV sterilization stages in zebrafish facilities are constructed to deliver 30 to 50,000 μWsec/cm², which is effective against several bacteria and algae. Studies investigating the efficacy of UV treatment specifically against P. neurophilia have not been reported. Published doses that are effective against other microsporidian spores vary from 6 to 19,000 μWsec/cm² for Encephalitozoon spp.^10,20 to 283,500 μWsec/cm² for Loma salmonae.³ ZIRC UV sterilizers deliver a minimum of 45,000 μWsec/cm² (90% UV transmittance, 200 gal/min flow rate, and 80% lamp power). Our detection of P. neurophilia infections in prefiltration sentinel animals but not postfiltration and UV sentinels in both ZIRC sentinels and fish submitted by other facilities to the ZIRC pathology service is evidence of the efficacy of UV sterilization. In addition, large amounts of particulate matter passing through a filtration system potentially could shield microsporidian spores from the effect of the UV light. We therefore recommend sufficient particle filtration before exposure to UV irradiation.

Third, sick and old fish may be a reservoir for infective microsporidian spores. Removing them from the population will decrease the potential transmission to other fish through cannibalism or spawning. Once microsporidiosis is diagnosed in prefiltration sentinel animals, targeted sampling of moribund fish and random sampling of fish in tank populations should be used to identify infected stocks and establish the prevalence of the pathogen within the facility.

Fourth, given the potential for transmission of microsporidiosis during spawning, fish that are used for crossing should not be spawned with fish derived from more than one other tank. In addition, a separate tank should be used for spawning and then cleaned according to disinfecting protocols. In August 2006, we modified the ZIRC breeding strategy such that wildtype fish used for crossing are dedicated to a particular stock and do not rejoin their siblings in a main tank of wildtype fish. This practice resulted in an immediate reduction in the diagnosis of microsporidiosis in the sentinel fish (ZIRC Animal Health Report, http://zebrafish.org/zirc/documents/health_report.php).

Fifth, our experiment involving seeding a noninfected tank with debris from a known infected tank demonstrates the presence of infective spores in tank debris, which may be a result of either spawning in the tank or fish death, with subsequent spore release as the carcass breaks down. Given the risk of transmission through exposure to tank debris, a clean autoclaved scrubber should be used for cleaning, and a clean autoclaved net should be used for every tank. In addition, we recommend that known infected tanks should be cleaned last, after all presumed noninfected tanks have been cleaned. Between tank uses, visible debris should be removed from the tank surface and the tank put into a bleach bath before being washed and rinsed. A bleach bath of at least 100 ppm chlorine, pH 7, for 10 min will kill 99% of P. neurophilia spores.⁸

Finally, once a tank of fish has been identified as positive for microsporidiosis, efforts should be made to spawn and replace the stock as soon as possible. Our data show that the infection spreads rapidly through a tank population. While screening a population of TU zebrafish that were from the same breeding event but reared to adults in 3 separate 20-gal tanks, we detected P. neurophilia infections in histologic sections and by PCR in fish from only 1 of the 3 tanks. By sampling 50 fish from each tank of 250, we would have detected 6% or greater prevalence of microsporidiosis with 95% confidence, according to the following formula:^1,29

Here n is sample size, p is probability of finding at least one case, d is number of affected animals, and N is population size. Considering the size of the stock and mixing of the embryos, larval, and juvenile fish, the data suggest that infective material or fish must have been present at a low enough level to be apportioned into only one of the 20-gal tanks. Therefore, our data indicate that the infection spread from this low level to 77% of the sampled population at 370 dpf and 90% of the sampled population at 517 dpf.

Studies documenting the progression of P. neurophilia infection in zebrafish are not reported. However, given the tropism of the parasite for neural tissue and the fact that we have not seen infection of ovary without concomitant infection of neural tissue, we surmise that infection with P. neurophilia starts in neural tissue and spreads to ovarian tissue at later stages. Similarly, in male zebrafish, the microsporidium may spread from the CNS to tissues around the testis during advanced stages of the infection. Therefore, the younger infected animals are spawned, the less likely they are to shed infective spores with the eggs or sperm. In addition, the amount of infective spores in fish that could be cannibalized or shed during spawning likely increases as the infection matures. PCR analysis can be performed on a percentage of the brood to determine whether infective material is present. This method has successfully been used to start and maintain a zebrafish facility that is SPF for P. neurophilia.¹⁴

In conclusion, P. neurophilia infections in laboratory zebrafish present a considerable concern for fish colony health and prospective research objectives. The infection was diagnosed in 74% of the zebrafish facilities that submitted samples to the ZIRC pathology service in 2010, and this prevalence is expected to spread as the use of zebrafish in research laboratories increases. Whether setting up a new zebrafish facility or trying to control infections in an existing fish population, the negative effect of microsporidiosis can be minimized, and hopefully eliminated, by initiating appropriate biosecurity measures and adjusting tank husbandry and breeding strategies.